Restriction endonucleases are key reagents for genetic engineering. Although more than two hundred specificities already exist, there is a need for more that recognize novel sequences. For instance, additional enzymes recognizing octanucleotide sequences would be useful for mapping and many disease genes contain mutations that could easily be detected if an appropriate restriction enzyme were available. Current biochemical screening methOds are inefficient, laborious and rarely successful. These assays find restriction enzymes in only one of four microorganisms examined and the rate of finding new specificities has dropped dramatically (only 3 in the last year). Current DNA sequencing efforts suggest that all bacteria possess at least one, and probably many more, Restriction- Modification (RM) systems. This proposal offers a new, computationally-based approach that screens the daily updates of GenBank for potential restriction enzyme genes and helps to identify those that may have new specificities. Candidates will be screened biochemically to test the predictions. New specificities will be fully characterized and offered for sale. PROPOSED COMMERCIAL APPLICATIONS This research will lead to the discovery of new restriction endonucleases with novel specificities. It will also identify clones and sequences for some previously uncloned enzymes. The new enzymes will be sold both to the Biotechnology industry and to the academic research community.